6
Imaging: Craniofacial Asymmetries
David C. Hatcher and Vidhya Venkateswaran
Introduction
A facial or craniofacial asymmetry is a significant sign or phenotype of deviant facial growth that has the potential to negatively impact facial appearance, occlusion, airway dimensions, and jaw function (Dadgar‐Yeganeh et al. 2021; Phi et al. 2022; Trypkova et al. 2000). The emphasis of this chapter is to provide methods and strategies to identify the underlying cause and quantify a craniofacial asymmetry thus leading to a precise diagnosis. A precise diagnosis can be folded into the clinical decision process addressing the therapy options and prognosis.
There is a wide range of congenital, developmental, and acquired conditions that can evolve into a craniofacial asymmetry. A precise diagnosis of an asymmetry can aid in predicting the clinical course of the condition. It is anticipated that therapy outcomes will be improved when tailored to the unique behavior and expression of underlying conditions over time.
The craniofacial structures are composed of a series of interconnected anatomic elements, including jaws, teeth, temporomandibular joints (TMJs), and muscles. Numerical modeling has shown that the functional interaction of the anatomic elements generates and propagates stresses and strains throughout craniofacial structures (Faulkner et al. 1987; Hatcher et al. 1986; Korioth and Hannam 1990). These stresses and resulting strains provide functional signals that can modulate tissue differentiation, growth, development, modeling, remodeling, and failure. Variability of neuromuscular interactions combined with the individual genetic and epigenetic influences on facial growth may contribute to a range of morphologic outcomes generated by apparently similar force stimuli.
The craniofacial structures contributing to asymmetry can be divided into two components – local and regional. Hard and soft tissues will be considered as local environments when they are related to the source of the asymmetry and as regional environments when they are adjacent to and biomechanically influenced by the local environment.
Asymmetry
Stress is an engineering term that expresses the force per unit area (Stress = Force/Area) while strain is the deformation of the structure during the application of force (Strain = change in size/original size). The functional interactions between components of anatomic environments lead to tissue stress and strain, subsequently inducing adaptive responses in the involved tissues. The adaptive tissue responses include changes in size and shape, structure, and 3D spatial position. The adaptive responses progress until an equilibrium is achieved between the functional demands and the capacity of the involved tissues (Carter and Beaupré 2007a, 2007b; Sommerfeldt and Rubin 2001).
Craniofacial asymmetry is a presenting sign for a wide variety of underlying biological causes. A selected group of diagnostic categories that are most relevant to the orthodontic community, primarily influencing the mandible and occlusion will be featured in this chapter. The local environment for most of these mandibular‐based asymmetries originate in the TMJ(s) and the regional environments extend to the remainder of the mandible, maxilla, skull base, occlusion, and airway.
Differential Diagnosis and Decision Tree for a Craniofacial Asymmetry
Differential diagnosis is a list of diagnostic possibilities for a given sign or symptom. In the case of a craniofacial asymmetry or phenotype (anatomic sign), it is diagnostically practical and efficient to organize the differential diagnosis into a decision tree. A decision tree presents decision points in a specific sequence that can be answered with discriminatory observations or applied tests and lead to a working radiographic diagnosis (Figure 6.1).
Imaging for Asymmetry
Choosing the Right Imaging Modality
Designing the optimal imaging study is a thoughtful process that begins with the decision to image and creating imaging objectives before selecting the imaging modality and imaging parameters. For example, in the case scenario of a clinically apparent facial asymmetry along with associated history, the clinician needs to decide if imaging will lead to clinically relevant findings. The best imaging studies are achieved when the clinician decides precisely what they desire the imaging study to reveal about the patient’s anatomy. Once imaging goals are known then the preferred imaging modality can be selected to achieve the desired diagnostic information while maintaining an acceptable cost and risk to the patient. Imaging methods have expanded and evolved to static and dynamic 2D and 3D methods. These methods include photo documentation, magnetic resonance imaging (MRI), and computed tomography (CT). In the case of craniofacial asymmetries, the field of view needs to encompass the involved local and regional environments. This field of view generally includes all of the maxilla, mandible, midface, TMJs, and temporal bone. The ideal imaging of skeletal asymmetry will allow for a metric quantification of the asymmetry. This can best be achieved with a three‐dimensional imaging method, such as cone‐beam CT (CBCT) and multidetector CT (MDCT). The resolution needs to be sufficient to assess the anatomic detail of the temporal bone, dentition, and TMJs.
Figure 6.1 DDx. This table illustrates a differential diagnosis and decision tree for common and selected disorders primarily of TMJ origin are responsible for the development of mandibular asymmetries. There are many other disorders leading to a mandibular asymmetry that are not portrayed on this table.
A CBCT is often preferred over MDCT because a CBCT has the most favorable value proposition including, cost, risk, spatial resolution, and acceptable tissue contrast. The usefulness of the imaging study is proportional to the appropriateness of the selected imaging modality, imaging quality, and the quality of the interpretation.
Volume Preparation for Analysis
Orientation: Image orientation is recommended prior to the imaging evaluation. When CBCT and MDCT scans are acquired, the scanner transfers its cartesian coordinate orientation to the produced image or voxel volume and this orientation is influenced by patient positioning. Post‐acquisition viewing software, such as Anatomage InVivo 6.0, is available to optimize the volume orientation, i.e. perform a Cartesian coordinate transformation to align it with specific anatomic structures. This orientation process allows for optimized visualization of the right and left sides of the skeleton to assess and compare the size, shape, and positional differences. The reorientation is performed with six degrees of freedom (X, Y, Z, Yaw, Pitch, and Roll) (Figure 6.2). Yaw is adjusted by viewing the skull base using an inferior view and aligning the coronal plane to two points on the skull base such as foramen spinosum. Pitch can be adjusted in the lateral view aligning the sagittal plane to porion and orbitale landmarks (Frankfurt plane). Roll is adjusted in the frontal view by aligning the axial plane it to the right and left orbital rims.
Similarly, evaluation of specific anatomy, such as the TMJs, requires optimization of the plane of section in order to accurately determine the size, shape, quality, and spatial relationships. The optimized TMJ views create oblique sagittal and oblique coronal views that are aligned perpendicular or parallel to the mediolateral long axis of the condyle.
Figure 6.2 Cartesian co‐ordinates (6 DOF): The orientation process allows for optimized visualization of the right and left sides of the skeleton to assess and compare the size, shape, and positional differences. The re‐orientation is performed with 6 DOF (X, Y, Z, Yaw, Pitch, and Roll). Yaw is adjusted viewing the skull base using an inferior view and aligning the coronal plane to 2 points on the skull base such as foramen spinosum. Pitch can be adjusted in the lateral view aligning the sagittal plane to porion and orbitale landmarks (Frankfurt plane). Roll is adjusted in the frontal view by aligning the axial plane it to the right and left orbital rims.
Analysis
CBCT volumes offer unique visualization and analysis opportunities when applied to the diagnosis and quantification of craniofacial skeletal asymmetries. Most of the encountered asymmetries display dimensional differences in size, shape, and spatial orientation when the right and left halves of the face are mirrored about the mid‐sagittal plane. The local and regional expression of a disorder are important to identify, visualize, and quantify in order to establish a baseline anatomy for future comparison and to arrive at a final radiographic diagnosis.
The initial step in craniofacial analysis is three‐dimensional landmarking of key anatomic features resulting in 3D coordinates for each landmark. The landmarking coordinates can be applied to various analysis strategies including generalized procrustes analysis (GPA), euclidean distance matrix analysis (EDMA), and principal component analysis (PCA).
Generalized Procrustes Analysis
GPA uses a 3D matrix of coordinate values for all of the local and regional anatomical landmarks of various individuals under study. As a first step, the shape of the structure under study can be mapped after removing confounding attributes of size, location, and rotation. This coordinate matrix of multiple individuals can then be compared to determine shape differences independent of size, location, and spatial orientation. For each patient under study, a landmark centroid is calculated using the coordinates values from the landmark matrix. Shape comparisons can be achieved by aligning all data sets onto the same coordinate system. This is achieved by superimposing the centroids, translating, rotating, and scaling the data matrices to minimize the distance between the corresponding landmarks. The aggregated cloud of data points for each landmark can be averaged to determine the mean location for each landmark. Subsequent Procrustes analysis can be performed to compare the average landmark locations with any given subject to help identify deviant anatomy.
Euclidian Distance Matrix Analysis (EDMA)
Euclidean distance involves calculating the distance between any two landmarks in 2D or 3D coordinate space using the Pythagorean theorem. The EDM for relevant anatomic elements associated with an asymmetry can be created for all components, for example, the length of the condylar process.
Principal Component Analysis
This is a method to reduce the dimensionality of data while retaining its primary characteristics. PCA isolates the landmarks that have the most significant influence of the data variance. PCA analysis includes computing the covariance matrix and eigenvectors and eigenvalues. Covariance measures how much two variables change together. Eigenvectors show the direction of the spread of data. Eigenvalues show the magnitude of the spread. Once the principal components are formed, PC1 will represent the variable with the greatest variance while PC2 will represent the variable with second most variance and so on. The greater the variance of a landmark location, the heavier its influence it has on the PCA analysis. Thus, the principal components with the greatest effect can be retained in the analysis, eliminating the others. Other variables with less influence can be eliminated in the subsequent analysis.
Examples of Asymmetrical Individuals
There are many conditions responsible for craniofacial asymmetry (Rathi and Hatcher 2019; Tamimi et al. 2017
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
